Machine learning and IoT-based cardiac arrhythmia diagnosis using statistical and dynamic features of ECG

Cardiac arrhythmia is a life-threatening disease which causes severe health problems in patients. A timely diagnosis of arrhythmia diseases will be useful to save the lives. Internet of Things (IoT) assures to modernize the health-care sector through continuous, remote and noninvasive monitoring of cardiac arrhythmia diseases. An IoT platform for prediction of cardiovascular disease using an IoT-enabled ECG telemetry system acquires the ECG signal, processes the ECG signal and alerts physician for an emergency. It is helpful for the physician to analyze the heart disease as early and accurate. We are developing an IoT-enabled ECG monitoring system to analyze the ECG signal. The statistical features of raw ECG signal are calculated. The ECG signal is analyzed using Pan Tompkins QRS detection algorithm for obtaining the dynamic features of the ECG signal. The system is used to find the RR intervals from ECG signal to capture heart rate variability features. The statistical and dynamic features are then applied to the classification process to classify the cardiac arrhythmia disease. People can check their cardiac condition by the acquisition of ECG signal even in their home. The size of the system is small, and it requires less maintenance and operational cost. It is helpful for the physician to analyze the heart disease as easily and accurately.